Using Artificial Intelligence & simulations to optimise traffic networks

Traffic lights in UK cities are about to get a brain transplant, swapping fixed timers for artificial intelligence that learns and adapts in real time. Current traffic signal systems operate on rigid schedules or simple sensor triggers, unable to respond to sudden congestion from accidents, events, or rush-hour surges. This project brings together a startup, a transport technology spin-out, and an industry veteran to build AI-driven control software that can simulate thousands of traffic scenarios and choose the optimal signal pattern for each moment. The team will integrate this system with existing urban traffic management hardware, then test and deploy it on real streets. If successful, the technology could cut average journey times, reduce idling at junctions, and lower vehicle emissions without costly road widening or new infrastructure. Drivers would spend less time stopped at red lights; delivery fleets and buses would run more predictably. The system could also adapt to temporary disruptions—a broken-down lorry, a road closure for a parade—without human intervention. For city planners, this means getting more capacity out of roads they already have, using software rather than concrete.